Determination of Total Phenolic Content and NIR-Chemometrics Classification Model of Queen and Local Varieties of Soursop (Annonamuricata L.) Leaf Powder

Abstract

The leaves of soursop (Annonamuricata L.) are commonly used for health because of their antioxidant activity from its highest phytochemical content, namely phenolic compound, which is influenced by the varieties of this plant. In Indonesia, there are two soursop varieties, namely ‘queen’ and ‘local’ varieties which are difficult to determine morphologically. The aim of this study was to determine the total phenolic content of soursop leaves of both varieties and to establish a classification model of NIR spectroscopy combined with chemometrics for the identification of the varieties of soursop leaves. After the soursop leaves were dried and grinded, they were then scanned to obtain the spectra of NIR spectroscopy. NIR spectras were combined with chemometrics to classify the varieties of the soursop. The classification models used were Linear Discriminant Analysis (LDA), Support Vector Machines (SVM) and Soft Independent Modelling of Class Analogies (SIMCA). Total phenolic content of the soursop leaves was determined by UV-Vis spectroscopy using Folin-Ciocalteau reagent and gallic acid as reference. The result showed that the local variety had higher total phenolic content than the queen variety. NIR spectroscopy combined with chemometrics was able to classify the varieties of soursop leaves with 100% accuracy using LDA and SVM.